Mapping tropical dry forest succession with chris - Proba hyperspectral images using non-parametric decissional trees
(2011) 34th International Symposium on Remote Sensing of Environment - The GEOSS Era: Towards Operational Environmental Monitoring- Abstract
Information on the distribution, level of fragmentation, extent and ecosystem health of tropical forests is key for sustainable landscape management. Therefore, information generated from different types of multispectral and hyperspectral sensors is key to monitoring programs that aimed to link ecosystems dynamics to ecosystem services such as carbon sequestration. This is especially significant for emerging tropical secondary forests which may play key roles on programs aimed to mitigate the effects of climate change. In this context, the aim of this presentation is to evaluate the accuracy of Chris/Proba and non - parametric classification as tools to monitor the ecological succession processes in tropical dry forests in Brazil... (More)
Information on the distribution, level of fragmentation, extent and ecosystem health of tropical forests is key for sustainable landscape management. Therefore, information generated from different types of multispectral and hyperspectral sensors is key to monitoring programs that aimed to link ecosystems dynamics to ecosystem services such as carbon sequestration. This is especially significant for emerging tropical secondary forests which may play key roles on programs aimed to mitigate the effects of climate change. In this context, the aim of this presentation is to evaluate the accuracy of Chris/Proba and non - parametric classification as tools to monitor the ecological succession processes in tropical dry forests in Brazil (Parque Estadual da Mata Seca, Minas Gerais). Our results from the non - parametric approach were separated into an age class map indicating the probably that a given pixel belongs to a forest succession class.
(Less)
- author
- Garcia-Millan, V. E. LU and Sanchez-Azofeifa, G. A.
- publishing date
- 2011-12-01
- type
- Contribution to conference
- publication status
- published
- subject
- keywords
- Chris proba, Hyperspectral, Non - parametric classification, Succession, Tropical dry forest
- pages
- 4 pages
- conference name
- 34th International Symposium on Remote Sensing of Environment - The GEOSS Era: Towards Operational Environmental Monitoring
- conference location
- Sydney, NSW, Australia
- conference dates
- 2011-04-10 - 2011-04-15
- external identifiers
-
- scopus:84879748169
- language
- English
- LU publication?
- no
- id
- 5fbe8b47-9f79-4b7c-90a5-f2110524e38f
- date added to LUP
- 2019-06-12 12:10:37
- date last changed
- 2022-01-31 21:41:36
@misc{5fbe8b47-9f79-4b7c-90a5-f2110524e38f,
abstract = {{<p>Information on the distribution, level of fragmentation, extent and ecosystem health of tropical forests is key for sustainable landscape management. Therefore, information generated from different types of multispectral and hyperspectral sensors is key to monitoring programs that aimed to link ecosystems dynamics to ecosystem services such as carbon sequestration. This is especially significant for emerging tropical secondary forests which may play key roles on programs aimed to mitigate the effects of climate change. In this context, the aim of this presentation is to evaluate the accuracy of Chris/Proba and non - parametric classification as tools to monitor the ecological succession processes in tropical dry forests in Brazil (Parque Estadual da Mata Seca, Minas Gerais). Our results from the non - parametric approach were separated into an age class map indicating the probably that a given pixel belongs to a forest succession class.</p>}},
author = {{Garcia-Millan, V. E. and Sanchez-Azofeifa, G. A.}},
keywords = {{Chris proba; Hyperspectral; Non - parametric classification; Succession; Tropical dry forest}},
language = {{eng}},
month = {{12}},
title = {{Mapping tropical dry forest succession with chris - Proba hyperspectral images using non-parametric decissional trees}},
year = {{2011}},
}